The last parts for the carbon microphone arrived yesterday! And once I got back to the Lair, I set about adding them to the circuit and building out the final version. Here’s a quick sample recording I’ll talk about more later.

I’m starting with the previously-discussed circuit, now taken out of the test harness and reassembled on one of the small breadboards. The new isolation transformer on the left – basically the outputs of the original circuit are attached to one side, and new outputs picked up off the other. This serves two purposes: first, it eliminates some kinds of hum noise if they start to crop up, and second, combined with 2K of resistance on the other side, brings the line-level(ish) output down to microphone level.

(These are standards which matter in a studio and … not many other places. XD )

The zig-zag in the resistors doesn’t serve any function purpose other than fitting into a smaller space – ideally, I guess I wouldn’t’ve been making needed values out of collections of other values? But I had what I had.

Signal flow is right to left in this photo. Underneath, at the top and bottom of the board, I’ve built wire rails to connect the components. That probably means I’m not really using the breadboard entirely as intended? I don’t even know. It holds everything in place and that’s definitely what I intended. XD

To house all this, I’m using an “experimenter’s case,” which is basically old-radio-speak for a metal box. XD It has soft metal on two sides, easily drillable and workable, and a hard case. I’m using one side for input, the power lamp, and the on/off switch; the other side is for output, or, as it turned out, outputs.

The carbon element (in the can) is connected to the driver/amplifier circuit via a 1/4″ TRS phone jack – like an old large headphone jack – with the two leads to the carbon element being on tip and ring, and the shielding ground being on the sleeve. (Tip, Ring, Sleeve: T R S.) That socket is on the left in the above photo; the middle component is a small LED, to indicate power on/off, and the right is a BIG CHUNKY POWER SWITCH. I love big chunky power switches. CHONK

For output, I quickly realised that I could have both balanced XLR output at microphone level, and line-level output on a phone plug, if I could find a way to isolate the chassis ground from the phone socket’s sleeve connector.

Normally, both being grounds on the same circuit, they’re connected automatically. Finding one that isn’t already connected is actively difficult! But careful use of electric tape did the job; I drilled the mounting hole larger than it needed to be, and basically lined anyplace the case and the socket would touch. Isolation achieved!

If you look for the blue and white wires, you can see where the TS (mono) phone plug is tapping the raw (line-level) amplified mic signal, just before it’s fed into the isolation transformer.

The transformer is really pretty optional – powered carbon circuit signals are pretty high as microphone signals go, and as I mentioned above, we’re actually reducing that signal to create the balanced XLR output on the other side of the transformer. But it’s nice to have the option of using line level, since it already exists. That’s what built-in sound inputs like on your laptop want, too, so there’s a point to it.

And here’s the whole driver/amplifier circuit, with a battery holder made of velcro.

Is that cheating? Holding the battery down with velcro, I mean. totally cheating I’m hoping it works out – I didn’t have a 9v battery case and it seemed excessive to try ordering one.

That LED power indicator? It’s warm white, left over from another project. I was planning on putting in your typical red LED, but realised that if they’d had a power indicator on one of these in 1932 or whenever, it most certainly would’ve been a little incandescent bulb, and it may and may not have had a colour lens. So I went with warm white, because period accuracy! Sort of.

The neat thing about the way this circuit works – and all carbon microphone driver circuits work – is how it points you right at vacuum tubes, and from there transistors, conceptually. It really, really does.

See, in tubes and transistors – which are both signal amplifiers – the input signal is used to create an amplified copy by controlling how much raw input power is let through, from another source. That’s why tubes were called “valves” originally; it’s because they are valves, electrically controlled, and regulating the flow of electricity from an input, just like the valve on your faucet controls the flow of water from the plumbing.

In this case, exactly the same thing happens yet again. But the input signal is sound pressure (how loud the sound is), which is controlling how much electricity is let through from the battery. And those changes in sound pressure – and therefore electrical flow – make the electrical copy of the sound waves.

Neat, huh?

Anyway, that’s the inside. Let’s look at the case!

I really like how chunky and primitive it looks. This is an old experimenter’s case; I’ve had a box of random cases in which I can build things for a while, and I don’t even know where I got this one, or when. If you saw it on the set of a 1950s television SF show, nobody would give it a second glance.

Always document your builds! You never know what might confuse people later. And by people, I mean yourself, after you’ve come down from the science-related memetic disorder high. I want at least the theoretical possibility of using this amp with other carbon elements, so writing down how the interface works is pretty important!

Except for the glare from the power light, I think this would be the Radio Shack Catalogue photo from, say, 1975:

Good, Better, or Best? Probably “Good.” It is just carbon, after all!

Or maybe this is the catalogue shot? Not sure.

Finally, here’s a test recording I made, using both outputs (phone/line level and XLR/balanced mic level) at once, hooked up to two different inputs on my board. I put both recordings in the same mp3; one’s on the left channel, the other’s on the right. The two tracks should be pretty much identical – being the same signal picked up at two different places on the board – and I wanted to see if that actually happened. Fortunately, it did!

Well, eventually it did. This is actually the second time I tried this, because the first time, I discovered that I’d managed to wire the two outputs up as electrical inverses of each other. Playing the two tracks back at the same time resulted in massive waveform cancellation. Which was hilarious, but also a good indicator; they wouldn’t’ve cancelled so well if they weren’t really similar. 😀


So that’s about it for this project! I’ll most likely do something to the ring to control the elastic better. And I’ll probably build a case for the whole kit, like I’ve done before – mics should have cases for protection! – but that’s a separate project.

This has been such a fun build, you have no idea. If you have any interest in this kind of DIY audio, I totally recommend this as a fun, easy project. Particularly if you don’t have studio gear, because you can look up the line level part of the output to damn near anything (including a PA system, I might mention) and it’ll work.

As always, more and bigger pictures on my Flickr account. And if you’re out of work, that’s a great time to listen to the new (NSFW lyrics) single! It’s awesome.